1. Field of the Invention
The present invention relates to a mask for use in exposing a photosensitive substrate with a pattern designed for devices, such as a semiconductor including an IC, an LSI, etc., a liquid crystal panel, a magnetic head and a CCD (image sensor), and an exposure method and an exposure apparatus which use the mask, and more particularly, relates to a mask, an exposure method and an exposure apparatus which are adapted for manufacturing such devices to a high degree of integration.
2. Description of Related Art
Heretofore, in manufacturing an IC, an LSI, a liquid crystal element, etc., by photolithography techniques, a projection exposure apparatus is used which performs an exposure by projecting through a projection optical system a pattern of a photomask or a reticle (hereinafter referred to generally as xe2x80x9cmaskxe2x80x9d) onto a photosensitive substrate, such as a wafer or a glass plate, which is coated with a photoresist or the like.
In recent years, the degree of integration of devices, such as an IC, an LSI and a liquid crystal element, is increasing more and more. As one of demands for minute and fine working of semiconductor wafers according to such an increase of the degree of integration, a device pattern is required to be more finely and minutely formed, i.e., to have a higher resolution.
For such a requirement, the projection exposure technique, which plays a main role in the art of accomplishing minute work, is being developed these days so as to form a pattern image of a line width not greater than 0.5 xcexcm over a wider range.
FIG. 21 is a schematic diagram showing the arrangement of a conventional exposure apparatus.
In FIG. 21, there are illustrated an excimer laser light source 171, an illumination optical system 172, illumination light 173, a mask 174, object-side exposure light 175, a projection optical system 176, image-side exposure light 177, a photosensitive substrate (wafer) 178, and a substrate stage 179 arranged to hold the photosensitive substrate 178.
In the conventional exposure apparatus shown in FIG. 21, a laser beam emitted from the excimer laser light source 171 is led to the illumination optical system 172. At the illumination optical system 172, the laser beam is converted into the illumination light 173 having a predetermined light intensity distribution, a predetermined luminous distribution, etc., and the illumination light 173 is then made incident on the mask 174. On the mask 174, a circuit pattern which is to be formed on the photosensitive substrate 178 is beforehand formed with chromium or the like in a predetermined magnified size. The illumination light 173, passing through the mask 174, is diffracted by the circuit pattern to be converted into the object-side exposure light 175. The projection optical system 176 converts the object-side exposure light 175 into the image-side exposure light 177 to image the circuit pattern on the photosensitive substrate 178 at a predetermined magnification with sufficiently small aberrations. As shown in an enlarged view at the lower portion of FIG. 21, the image-side exposure light 177 converges on the photosensitive substrate 178 at a predetermined NA (numerical aperture=sin xcex8) to be imaged there. To have the circuit pattern formed in a plurality of shot areas on the photosensitive substrate 178, the substrate stage 179 is arranged to be movable stepwise to vary the relative positions of the photosensitive substrate 178 and the projection optical system 176.
In the above projection exposure apparatus using an excimer laser, which is currently widely used, however, it is difficult to form a pattern image of a line width not greater than 0.15 xcexcm.
The reason for this difficulty is explained below. The resolution of the projection optical system is limited by a trade-off between the optical resolution and the depth of focus due to the wavelength of the exposure light. The resolution R and the depth of focus DOF in the resolving pattern by the projection exposure apparatus can be expressed by the following Rayleigh""s formulas (a1) and (a2):                     R        =                              k            1                    ⁢                      λ            NA                                              (a1)                                DOF        =                              k            2                    ⁢                      λ                          NA              2                                                          (a2)            
where xcex is the wavelength of the exposure light, NA is a numerical aperture indicative of the brightness of the optical system, k1 and k2 are constants which are determined by the developing process characteristic, etc., of the photosensitive substrate and are normally between 0.5 and 0.7.
According to the formulas (a1) and (a2), in order to make the value of the resolution R smaller for a higher degree of resolution, it is necessary either to make the wavelength xcex smaller for a shorter wavelength or to make the NA larger for a higher degree of brightness. At the same time, however, the depth of focus DOF required for a necessary performance of the projection optical system must be kept at least at a certain value. This requirement imposes some limitation on the increase of the NA. As a result, the shortening of the wavelength is considered a sole solving method.
The attempt to shorten the wavelength, however, encounters several serious problems other than the problem related to the above formulas. The most serious problem lies in that it becomes hardly impossible to find any optical material usable for the projection optical system. An optical system which is actually mountable on the exposure apparatus as the current projection optical system in view of the amount of aberration, the precision of working, the controllability, etc., is one including a refractive system, i.e., a lens. Almost all optical materials used for lenses have transmission factors near xe2x80x9c0xe2x80x9d in the short wavelength region, i.e., in the far ultraviolet region. Although there are a fused quartz material, etc., as an optical material which is manufactured by a special manufacturing method for an exposure apparatus, the transmission factor of the fused quartz also abruptly drops for the wavelength not greater than 193 nm. It is thus extremely difficult to develop any optical material practically usable for an exposure wavelength not greater than 150 nm required for a pattern of a line width not greater than 0.15 xcexcm, because the optical material is required to satisfy a plurality of conditions relative to durability, uniform refractive index, optical strain, workability, etc., in addition to the transmission factor.
The above conventional projection exposure method necessitates the shortening of the wavelength for performing a pattern exposure depending on the formulas (a1) and (a2), and, therefore, causes a problem in that there exists no usable optical material, so that it is impossible to realize an exposure for a pattern image of a line width not greater than 0.15 xcexcm.
There is another exposure method called a two-light-flux interference exposure method. FIG. 22 is a schematic diagram for explaining the two-light-flux interference exposure method. According to the two-light-flux interference exposure method, coherent light emitted from a laser light source 71 is divided by a half-mirror 72 into two light fluxes. Mirrors 73a and 73b are arranged to deflect the two light fluxes respectively at some angles to cause the two light fluxes to join together on a photosensitive substrate 74 in such a way as to form interference fringes there. Then, the photosensitive substrate 74 is exposed according to a distribution of light intensity made by the interference fringes, so that a periodic pattern is formed according to the distribution of light intensity.
The resolution R obtained by the two-light-flux interference exposure method is expressed by the following formula (a3), where the resolution R is assumed to be the width of each of lines and spaces (LandS), i.e., the width of each of bright and dark bands of the interference fringes, e represents the angle of incidence on the photosensitive substrate 74 of each of the two light fluxes 71a and 71b, and NA sin xcex8.                                                         R              =                              xe2x80x83                            ⁢                              λ                                  4                  ⁢                                      xe2x80x83                                    ⁢                  sin                  ⁢                                      xe2x80x83                                    ⁢                  θ                                                                                                        =                              xe2x80x83                            ⁢                              λ                                  4                  ⁢                                      xe2x80x83                                    ⁢                  NA                                                                                                        =                              xe2x80x83                            ⁢                              0.25                ⁢                                  xe2x80x83                                ⁢                                  λ                  NA                                                                                        (a3)            
As is understandable from the formulas (a3) and (a1), the constant k1 becomes 0.25 (k1=0.25) according to the two-light-flux interference exposure method. Considering that the value of the constant k1 in the case of the conventional projection exposure method is between 0.5 and 0.7, the resolution obtainable by the two-light-flux interference exposure method is more than two times as high as the resolution obtainable by the conventional projection exposure method. According to the two-light-flux interference exposure method, assuming that xcex is 0.248 xcexcm and NA is 0.6, for example, the resolution R becomes 0.10 xcexcm.
However, the two-light-flux interference exposure method presents a fundamental problem in that it is impossible to perform an exposure with a circuit pattern composed of diverse shapes like semiconductor element patterns. In other words, according to the two-light-flux interference exposure method, only such a pattern that has a uniform pitch over the whole area where the two light fluxes are joined together is substantially available for exposure. In order to improve such a problem, a method is conceivable, for example, in which a diaphragm is set near the surface of the photosensitive substrate 74 to limit an area to be exposed and an exposure is performed with the pattern of several lines and spaces required for a specific pattern, while changing the exposure direction and moving the exposure position. However, in a case where such a method is executed, at the same time that a small diaphragm is provided to limit the exposure area to a sufficiently small area, a distribution of light intensity would be changed due to the diffraction of individual light fluxes themselves, so that it becomes impossible to form effective interference fringes.
While, according to the two-light-flux interference exposure method, the resolution corresponding to the line width of not greater than 0.15 xcexcm can be accomplished, the resolvable pattern is limited to a repetitive pattern of uniform pitch, so that it is impossible to perform an exposure with sufficiently various kinds of circuit patterns required for the practical semiconductor devices.
Further it is widely known that the Levenson-type reticle is used in order to realize the two-light-flux interference exposure method using a projection exposure apparatus. However, there is a problem that it is difficult to make such a Levenson-type reticle that has sufficiently various kinds of circuit patterns required for the practical semiconductor devices. In addition, in the case of usage of the Levenson-type reticle, the oblique incidence illumination is inferior in contrast, and there is such a limitation that the inclination of an incident light beam has no freedom (allowing only one angle).
To solve the above problems, a multiple exposure method using both a two-light-flux interference exposure and a projection exposure has been proposed in Japanese Patent Application Laid open No. 11-143085.
Next, the multiple exposure method proposed in Japanese Patent Application Laid open No. 11-143085 will be explained taking an example thereof.
One of circuit patterns which are obtainable according to the multiple exposure method is a so-called gate pattern as shown in FIG. 12. The gate pattern has such a characteristic that the line width in the horizontal direction is 0.1 xcexcm while the line width in the vertical direction is 0.2 xcexcm. For such a pattern that a high resolution is required only in a one-dimensional direction, an exposure in the two-light-flux interference exposure step can be performed only in the one-dimensional direction requiring a high resolution.
As examples of exposure apparatuses capable of attaining the two-light-flux interference exposure, there are an exposure apparatus shown in FIG. 22 and a projection exposure apparatus shown in FIG. 23, which is capable of selecting masks and illumination methods as shown in FIGS. 24(A) and 24(B) and FIG. 25.
It is to be noted that the term xe2x80x9cprojection exposurexe2x80x9d in the current description is used in the narrow sense. Specifically, the projection exposure in the broad sense includes an exposure system in which the two-light-flux interference exposure is substantially attained by adaptively selecting and adjusting the mask and the illumination method as shown in FIG. 23. On the other hand, in the projection exposure in the narrow sense, an exposure with an arbitrary pattern formed on the mask is performed by using convergent light fluxes incident on the image plane at various angles including two parallel light fluxes.
First, an explanation is made about the exposure apparatus shown in FIG. 22. In the exposure apparatus, the line width of an interference pattern obtainable by exposure is expressed by the above formula (a3) according to the angle of incidence xcex8 of each of two light fluxes which are joined together. Here, the angle xcex8 and the NA are in the relation of xe2x80x9cNA=sin xcex8xe2x80x9d. Since the angle e is settable to an arbitrary value by adjusting the angle of the mirror 73a or 73b, it is possible to greatly decrease the line width of an interference pattern by setting the angle xcex8 to a larger value. For example, in a case where a KrF laser of wavelength 248 nm is used as a light source, assuming that the angle xcex8 is set to 38 degrees, the possible line width of an interference pattern becomes 0.1 xcexcm. In this instance, the NA (=sin xcex8) is 0.62. If the angle xcex8 is set to a larger value, a higher resolution can be obtained.
Next, an explanation is made about the exposure apparatus shown in FIG. 23. The exposure apparatus shown in FIG. 23 is provided with a projection optical system, which is conventionally available and is entirely composed of, for example, refractive systems, and may have the NA not less than 0.6 for the wavelength 248 nm. In FIG. 23, the illustration includes a mask 81, object-side exposure light 82, a projection optical system 83, an aperture stop 84, image-side exposure light 85, a photosensitive substrate 86, and a schematic view 87 showing the positions of light fluxes at the pupil surface of the projection optical system 83.
In the schematic diagram of FIG. 23, which shows the state of the exposure apparatus where the two-light-flux interference exposure is being performed, each of the object-side exposure light 82 and the image-side exposure light 85 is composed of two parallel light fluxes.
In such an ordinary projection exposure apparatus, the two-light-flux interference exposure can be performed by setting the mask 81 and the illumination method as shown in FIGS. 24(A), 24(B) and 25, which will be explained below as three examples.
In the example shown in FIG. 24(A), there is illustrated a Levenson-type phase shifting mask 90, in which the pitch P0 of chromium-made light-blocking parts 91 is expressed by the following formula (a4) and the pitch P0S of phase shifters 92 is expressed by the following formula (a5):                               P          0                =                              P            M                    =                                                    2                ⁢                R                            M                        =                          λ                              2                ⁢                                  xe2x80x83                                ⁢                                  NA                  ·                  M                                                                                        (a4)                                          P                      0            ⁢            S                          =                              2            ⁢                          P              0                                =                      λ                          NA              ·              M                                                          (a5)            
where P is the pitch of interference fringes, M is the magnification of the projection optical system 83, xcex is the wavelength, and NA is the image-side numerical aperture of the projection optical system 83.
Further, in the example shown in FIG. 24(B), there is illustrated a phase shifting mask 93 of the shifter-edge type having no chromium-made light-blocking part, in which the pitch P0S of phase shifters 92 is expressed by the above formula (a5) as in the case of the Levenson type.
In each of the phase shifting masks 90 and 93 shown in FIGS. 24(A) and 24(B), when the illumination is performed with parallel light fluxes which are made vertically incident (coherence factor "sgr"=0), vertically-transmitted light fluxes cancel each other and disappear because the phase difference between adjacent transmitted light fluxes is made to be xcfx80 by the phase shifters 92. Then, two light fluxes for xc2x1 first-order diffracted light, which strengthen each other, occur symmetrically with respect to the optical axis of the projection optical system. Further, second-order or higher-order diffracted light does not contribute to the image formation, owing to the limitation of the NA of the projection optical system.
In the example shown in FIG. 25, there is illustrated an ordinary chromium mask 101, in which the pitch P0 of chromium-made light-blocking parts is expressed by the following formula (a6), which is the same as the formula (a4):                               P          0                =                              P            M                    =                                                    2                ⁢                R                            M                        =                          λ                              2                ⁢                                  xe2x80x83                                ⁢                                  NA                  ·                  M                                                                                        (a6)            
In the case of the mask having no phase shifting structure, as shown in FIG. 25, illumination light is subjected to the oblique incidence illumination. In this case, incident light is composed of parallel light fluxes having an angle of incidence xcex80. Here, the angle of incidence xcex80 is expressed by the following formula (a7):
sin xcex80=Mxc2x7NAxe2x80x83xe2x80x83(a7)
With respect to such incident light, light having passed through the mask 101 is composed of two light fluxes, i.e., zero-order diffracted light which advances at the angle xcex80 relative to the optical axis of the projection optical system in the same way as the incident light and xe2x88x92first-order diffracted light which advances at the angle xe2x88x92xcex80 symmetrically to the zero-order diffracted light with respect to the optical axis of the projection optical system. Then, the two light fluxes pass through the projection optical system and contribute to the image formation.
The above explanation has been made about examples of the mask and the illumination method for performing the two-light-flux interference exposure using the ordinary projection exposure apparatus. With the above-described setting, a maximum area of the NA of the projection optical system can be utilized.
Subsequently, an explanation is made back about the exposure method. In the above example, the two-light-flux interference exposure step (FIG. 13(A)) is followed by the projection exposure step, which is executed to perform an exposure with a pattern shown in FIG. 13(B). In the upper portion of FIG. 13(B), there are shown amounts of exposure on the individual areas in the projection exposure step with regard to the positions of interference fringes formed by the two-light-flux interference exposure step. In the lower portion of FIG. 13(B), there are shown the amounts of exposure in the form of a map at the resolution of a pitch of 0.1 xcexcm.
Here, it is found that the line width of the pattern formed by the projection exposure step is 0.2 xcexcm, which is twice as large as that formed by two-light-flux interference exposure step.
Further, as one of methods for performing a projection exposure in which amounts of exposure differ with the individual areas, as described above, there is a method of using a mask having a plurality of steps of transmission factors, in which a mask part corresponding to the area indicated by xe2x80x9c1xe2x80x9d in FIG. 13(B) is assigned a transmission factor of T% and a mask part corresponding to the area indicated by xe2x80x9c2xe2x80x9d in FIG. 13(B) is assigned a transmission factor of 2T%. According to this method, the projection exposure step can be completed by one exposure. In this case, a ratio among the amounts of exposure obtained on the photosensitive substrate by the respective exposure steps is as follows: xe2x80x9ctwo-light-flux interference exposurexe2x80x9d:xe2x80x9cprojection exposure at the part of transmission factor T%xe2x80x9d:xe2x80x9cprojection exposure at the part of transmission factor 2T%xe2x80x9d=1:1:2.
Further, there is another method in which an exposure is performed twice at the same amount of exposure by using two masks having respective different patterns as shown in the upper and lower portions of FIG. 13(D).
Next, the formation of a lithography pattern by the combination of the two-light-flux interference exposure step and the projection exposure step is described. In the case of such a multiple exposure method, there is no developing step between the two-light-flux interference exposure step and the projection exposure step. Therefore, the amount of exposure obtained by one exposure step is added to the amount of exposure obtained by the other exposure step. Then, a total amount of exposure obtained by the addition causes a new exposure pattern (distribution of latent images or distribution of amounts of exposure) to be formed.
FIG. 13(C) illustrates the results of addition of the amounts of exposure obtained by the above two kinds of exposure steps. Particularly, in the lower portion of FIG. 13(C), a lithography pattern obtained by developing the exposure pattern shown at the upper portion of FIG. 13(C) is shown in a gray tone. Additionally stated, the exposure threshold value of the photosensitive substrate used in the above example is a value not less than xe2x80x9c1xe2x80x9d and less than xe2x80x9c2xe2x80x9d.
The lithography pattern shown in gray at the lower portion of FIG. 13(C) coincides with the gate pattern shown in FIG. 12, so that it is found that the above multiple exposure method enables the formation of such a complicated minute pattern.
As a result of diligent studies of the inventor of the present invention, it is found that the following problem arises in a case where the conventional half-tone mask is used as a half-tone mask provided with a light-transmitting part whose transmission factor is not 100%, as described by referring to the upper portion of FIG. 13(B).
FIG. 26 schematically shows the manner of the projection exposure in a case where an exposure with the above gate pattern is performed on the resist by using a half-tone mask.
In the upper portion of FIG. 26, there is illustrated a multivalued half-tone mask for use in the projection exposure step to form the gate pattern. In this case, three values of transmission factors are provided, i.e., 0%, 50% and 100%. In the middle portion of FIG. 26, there are shown the transmission factor, the amount of phase change of transmitted light and the complex amplitude transmission factor, which are obtained on a section B-Bxe2x80x2 of the half-tone mask.
The conventional half-tone mask is accompanied by the phase shifting effect, which aims at improving the resolution by varying the complex amplitude transmission factor according to the amount of phase change and adjusting the amplitude corresponding to an exposure pattern. Therefore, the amount of phase change of the conventional half-tone mask takes a value of, for example, xcfx80, as shown in FIG. 26.
If such a half-tone mask is used for the above-described multiple exposure, although the numerical values of the transmission factors are set to 0%, 50% and 100%, the complex amplitude transmission factors become 0%, xe2x88x9250% and 100%, so that there exists a portion in which the complex amplitude transmission factor becomes xe2x80x9c0xe2x80x9d at the boundary between 50% and 100%. Accordingly, the actual transmission factor, i.e., the amount of exposure, becomes xe2x80x9c0xe2x80x9d in the vicinity of the portion in which the complex amplitude transmission factor becomes xe2x80x9c0xe2x80x9d, as shown in the lower portion of FIG. 26. Such abnormality makes it impossible to obtain a desired, continuous multivalued half-tone mask required for the above multiple exposure.

It is an object of the invention to eliminate the problems arising when the above half-tone mask is used for the multiple exposure and to provide a mask and an exposure method using the mask, which make it possible to form on a wafer a pattern of line width not greater than 0.15 xcexcm even by means of the ordinary projection exposure apparatus, and which enable a high-integrated device to be readily manufactured.
To attain the above object, in accordance with a first aspect of the invention, there is provided a mask, which comprises a substrate, and a pattern having a transmission factor formed on the substrate by using a material, wherein an optical path length difference between light beams respectively passing the pattern and an area adjacent thereto is greater than       (          m      -              1        8              )    ⁢  λ
and less than             (              m        +                  1          8                    )        ⁢    λ    ,
where xcex is a wavelength of incident light, and m is an integer.
In accordance with a second aspect of the invention, there is provided a mask, which comprises a substrate, and a pattern having a reflection factor formed on the substrate by using a material, wherein an optical path length difference between light beams respectively passing the pattern and an area adjacent thereto is greater than       (          m      -              1        8              )    ⁢  λ
and less than       (          m      -              1        8              )    ⁢  λ
where xcex is a wavelength of incident light, and m is an integer.
In particular, the best value of the optical path length difference in each of the first and second aspects is mxc2x7xcex in principle.
Specifically, in the mask according to the first or second aspect, the wavelength xcex is in the following range:
150 nm less than xcex less than 440 nm.
In accordance with a third aspect of the invention, there is provided an exposure method, which comprises a step of performing an exposure by, while using the mask according to the first or second aspect, forming a pattern image of the mask on a photosensitive substrate.
In accordance with a fourth aspect of the invention, there is provided an exposure method, which comprises a projection exposure step of performing a projection exposure on a photosensitive substrate by using a pattern of the mask according to the first or second aspect, and a two-light-flux interference exposure step of exposing the photosensitive substrate with interference fringes (pattern) based on two-light-flux interference.
In accordance with a fifth aspect of the invention, there is provided an exposure apparatus, which has an exposure mode related to one or both of the exposure methods according to the third and fourth aspects.
In accordance with a sixth aspect of the invention, there is provided a device manufacturing method, which comprises a process of transferring a device pattern of a reticle to a wafer by using one of the exposure methods according to the third and fourth aspects, and a process of developing the wafer.